1. Field of the Invention
The present invention relates to a computer program product, system, and method for determining priorities for cached objects to order the transfer of modifications of cached objects based on measured network bandwidth.
2. Description of the Related Art
Wide Area Network (WAN) caching involves replicating data from a primary or source site to one or more secondary site to address WAN performance and network outage issues. The primary and secondary sites may provide network file server services, such as with Network Attached Storage (NAS) servers. With WAN caching, modifications to data cached at remote sites is pre-populated to remote caches at remote sites as soon as changes are made at the primary site for files cached at remote sites to make the data readily available at the remote site. With WAN caching, the data cached at the remote sites is available to users connecting to the remote sites even after disconnection between the remote cache site and the primary source site.
In a large enterprise environment, a large number of files at the primary site may be modified, which results in a large queue of modifications to replicate to the remote sites, which may consume a significant amount of network bandwidth. Further, in certain countries and regions, network bandwidth is very low, so WAN caching may experience bottlenecks during periods of low network bandwidth.
There is a need in the art for improved techniques for optimizing WAN caching data at remote secondary sites.